Transcript VLTI
Armando DOMICIANO de SOUZA Main collaborators: O. Chesneau (OCA, F), T. Driebe (MPIfR, D), K-.H. Hofmann (MPIfR, D), S. Kraus (MPIfR, D), A. Miroshnichenko (UT, US), K. Ohnaka (MPIfR, D), P. Stee (OCA, F), G. Weigelt (MPIfR, D) Plan Introduction The B[e] phenomenon Principles of optical/IR long baseline interferometry VLTI (MIDI and AMBER) observations of CPD-57 2874 VLTI-MIDI (visibilities, spectrum, modelling, comparison to other data) VLTI-AMBER (visibilities, modelling, phases) Comparison of VLTI-MIDI and VLTI-AMBER results The B[e] phenomenon (Lamers et al. 1998) 1. Strong Balmer emission lines. 2. Low excitation permitted emission lines of predominantly low ionization metals in the optical spectrum, e.g. Fe II. 3. Forbidden emission lines of [Fe II] and [O I] in the optical spectrum. 4. A strong near or mid-infrared excess due to hot circumstellar dust. Meilland The B[e] phenomenon (Lamers et al. 1998) Supergiants B[e] L*/Lsun > 104 Observations point towards asymmetrical stellar environments Need for direct measurements High angular resolution Zickgraf et al. (1985) Principles of optical/IR long baseline interferometry IOTA spectro-interferometry Bands J H K Weigelt et al. (2003) Principles of optical/IR long baseline interferometry Interference fringes Intensity map I(y,z,l) Weigelt et al. (2000) GI2T - g Cas Complex Visibilities V(u,v,l)= FT[I(u,v,l)] / FT[I(0,0,l)] Interferometry : the uv or Fourier plane ESO-VLTI Intensity distribution of the object at a given l Complex Visibility V(u,v,l) ASPRO - JMMC Fourier Transform u and v spatial frequency Bproj / l Partial uv coverage Models are needed to interpret the current interferometric observations Observations of B[e] stars with VLTI-MIDI and VLTI-AMBER Observational set-up: MIDI N band with R=30 (8-13 m) Unit Telescopes (UTs) 2 baselines AMBER K band with R=1200 (2.1-2.2 m) Unit Telescopes (UTs) 3 baselines (closure phase) Targets: GG Car MIDI Not well resolved (size < 10 mas) CPD-57 2874 MIDI and AMBER VLTI-MIDI observations MIDI Equivalent uniform disc model: V(l) = |2J1(z) / z| , where z = p qUD(l) Bproj / l VLTI-MIDI observations UD diameter versus Position Angle VLTI-MIDI : fit of V with gaussian-models Chromatic variation of the major axis FWHM 2a = 2a0 + K (l-l0) Gaussian circle 2a = (8.70.4) + (2.20.3) (l-8m) mas 2a = (13.50.2) + (0.40.2) (l-12m) mas Gaussian ellipse 2a = (10.10.7) + (2.60.4) (l-8m) mas Axial ratio 2b/2a = 0.76 0.08 Position angle PA = 144° 6° 2a = (15.30.7) + (0.50.2) (l-12m) mas Axial ratio 2b/2a = 0.80 0.06 Position angle PA = 143° 6° VLTI-MIDI spectrum Possible origin of this featureless spectrum around 10 m: Large grains ? Carbonaceous dust ? Free-free emission ? Additional opacity sources ? Modelling VLTI-MIDI observations Envelope of dust with spherical symmetry DUSTY code (Ivezic et al.) Stellar input parameters: distance = 2 kpc V = 10.1 Av = 5.9 V0 = 4.2 Teff = 20000 K log L/L = 5.6 R = 53R angular diameter Ø = 0.25 mas Spherical model (DUSTY code) : silicate with large grains Spherical model (DUSTY code) : silicate with large grains Spherical model (DUSTY code) : graphite with large grains Spherical model (DUSTY code) : graphite with large grains Dust close to the star ? SED can be reproduced by the spherical dust model, but not the visibilities inner dust radius is too large (~12 mas for silicates and ~60 mas for graphite) ! What is the origin of the mid-IR emission relatively close to the star measured with VLTI-MIDI ? Possibility to get dust closer to the star : Dense equatorial wind disk-like structure able to shield the disk material to allow molecules and dust to be formed near the hot central star (Kraus & Lamers 2003). Support for a non-spherical envelope • A spherical model does not seem to simultaneously fit the SED and VLTI-MIDI visibilities • Winds of sgB[e] have two components (e.g. Zickgraf et al. 1985) • Several sgB[e] show high intrinsic polarizations consistent with non-spherical dusty envelopes (e.g. Magalhães 1992) Zickgraf et al. (1985) Polarization PA versus VLTI-MIDI PA polarization <PA> = 45°3° U Data from Yudin & Evans (1998) B V Yudin & Evans (1998) Polarization perpendicular to disc: (45° 3°) + 90°= 135 3° N Ellipse orientation from MIDI : 143.5° 6° E VLTI-AMBER observations AMBER Equivalent uniform disc model: V(l) = |2J1(z) / z| , where z = p qUD(l) Bproj / l VLTI-AMBER observations UD diameter versus Position Angle VLTI-AMBER : fit of V with gaussian-models Chromatic variation of the major axis FWHM : 2a=2a0+K(l-l0)+C exp[-4ln2(l- lc)/l] Gaussian circle 2a = (2.460.01) + (5.20.1) (l-2.2m) mas Brg C = 0.640.08 mas ; l=1.6 0.2 pm Gaussian ellipse 2a = (3.900.03) + (7.30.2) (l-2.2m) mas Axial ratio 2b/2a = 0.56 0.01 Position angle PA = -2.9° 0.4° Brg C = 0.930.11 mas ; l=1.6 0.2 10-3 m closure phase (deg) VLTI-AMBER : closure phase VLTI Centrally-symmetric intensity distribution l (microns) Differential phase UT2-UT3 Differential phase Differential phase VLTI-AMBER : differential phases UT4-UT2 l (microns) UT3-UT4 VLTI No chromatic variation of object’s symmetry l (microns) Measured sizes of CPD-57 2874 AMBER N E MIDI FIN THE END FIM ENDE Interstellar polarization ? Stars within 2° of CPD-57 2874 (Heiles 2000) Stars with low and high polarizations have random PA Modelling VLTI-MIDI observations Inner radius Silicate rin=12 mas ~100R* ~ 24 AU Graphite rin=60 mas ~480R* ~ 120 AU Theory of (anisotropic) winds of massive stars Maeder & Desjacques (2001 A&A), Lamers & Pauldrach (1991 A&A), Maeder (1999 A&A), Langer et al. (1999 ApJ), etc Rapid rotation and Log L/L > 104 Star close to the -limit : bi-stable winds Lamers model geff-effect Eddington factor variable in latitude -effect von Zeipel effect: Maeder model van Boekel (2003) VLTI-VINCI Mass loss variable in latitude (opacity and gravity effect) : Car